Retirement—A Transition to a Healthier Lifestyle?

Evidence From a Large Australian Study

      Introduction

      Population aging is associated with a rising burden of non-communicable disease, profoundly impacting health policy and practice. Adopting and adhering to healthy lifestyles in middle or older age can protect against morbidity and mortality. Retirement brings opportunities to reconfigure habitual lifestyles and establish new routines. This study examines the longitudinal association between retirement and a range of lifestyle risk behaviors among a large population-based sample of Australian adults.

      Methods

      Study sample included working adults aged ≥45 years at baseline (2006–2009, N=23,478–26,895). Lifestyle behaviors, including smoking, alcohol use, physical activity, diet, sedentary behavior, and sleep, were measured at both baseline and follow-up (2010). Logistic regression models estimated the odds of having each risk factor at follow-up and multiple linear regression models calculated the change in the total number of risk factors, adjusted for baseline risk and other covariates. Sociodemographic characteristics and reasons for retirement were tested as potential effect modifiers.

      Results

      During the 3.3-year follow-up, about 11% of respondents retired. Retirement was associated significantly with reduced odds of smoking (AOR=0.74); physical inactivity (AOR=0.73); excessive sitting (AOR=0.34); and at-risk sleep patterns (AOR=0.82). There was no significant association between retirement and alcohol use or fruit and vegetable consumption. Change in the total number of lifestyle risk factors differed significantly by reason for retirement.

      Conclusions

      In a large population-based Australian cohort, retirement was associated with positive lifestyle changes. Health professionals and policymakers should consider developing special programs for retirees to capitalize on the healthy transitions through retirement.

      Introduction

      Many societies face the challenge of population aging. Globally, between 2000 and 2050, the proportion of adults older than 60 years is projected to double from 11% to 22%, and the number is expected to grow from 605 million to more than 2 billion.
      WHO
      10 Facts on Ageing and the Life Course.
      This demographic shift will lead to a rising disease burden of noncommunicable disease, with profound implications for health policy and practice.
      • Prince M.J.
      • Wu F.
      • Guo Y.
      • et al.
      The burden of disease in older people and implications for health policy and practice.
      Evidence suggests that adopting and adhering to healthy lifestyles in middle or older age can reduce risks for chronic disease and mortality.
      • Knoops K.T.
      • de Groot LC
      • Kromhout D
      • et al.
      Mediterranean diet, lifestyle factors, and 10-year mortality in elderly European men and women: the HALE project.
      • King D.E.
      • Mainous 3rd, A.G.
      • Geesey M.E.
      Turning back the clock: adopting a healthy lifestyle in middle age.
      Therefore, it is important to identify opportunities for lifestyle changes in older age to promote healthy aging.
      Transitioning out of the workforce represents one such critical moment for lifestyle modification. Retirement is associated with changes in time availability and flexibility, social contacts and networks, income, and financial security, all of which are related to lifestyle.
      • Kim J.E.
      • Moen P.
      Retirement transitions, gender, and psychological well-being: a life-course, ecological model.
      • Wang J.
      • Lin W.
      • Chang L.
      The relationships between retirement, social capital, and self-perceived health.
      Retirement may also allow individuals to rethink habitual behaviors and establish new routines.
      • Jonsson H.
      • Josephsson S.
      • Kielhofner G.
      Narratives and experience in an occupational transition: a longitudinal study of the retirement process.
      Therefore, transitioning into retirement could provide a window of opportunity for positive behavioral changes.
      Previous studies have explored the associations between retirement and health behaviors. A recent systematic review
      • Barnett I.
      • Guell C.
      • Ogilvie D.
      The experience of physical activity and the transition to retirement: a systematic review and integrative synthesis of qualitative and quantitative evidence.
      found a consistent increase in leisure-time physical activity after retirement, though change in total physical activity was less clear. Findings on retirement and alcohol consumption are mixed and suggest that the impact of retirement on drinking is complex and context specific.
      • Kuerbis A.
      • Sacco P.
      The impact of retirement on the drinking patterns of older adults: a review.
      • Zantinge E.M.
      • van den Berg M.
      • Smit H.A.
      • Picavet HSJ
      Retirement and a healthy lifestyle: opportunity or pitfall? A narrative review of the literature.
      Research on change in dietary behaviors and smoking after retirement is inconclusive.
      • Zantinge E.M.
      • van den Berg M.
      • Smit H.A.
      • Picavet HSJ
      Retirement and a healthy lifestyle: opportunity or pitfall? A narrative review of the literature.
      The literature regarding effect of retirement on sleep and sedentary behavior is limited. It appears that retirement may have a positive impact on sleep patterns,
      • Vahtera J.
      • Westerlund H.
      • Hall M.
      • et al.
      Effect of retirement on sleep disturbances: the GAZEL prospective cohort study.
      • Marquiae J.C.
      • Folkard S.
      • Ansiau D.
      • Tucker P.
      Effects of age, gender, and retirement on perceived sleep problems: results from the VISAT combined longitudinal and cross-sectional study.
      and some evidence suggests that retirement is associated with increased TV viewing.
      • Touvier M.
      • Bertrais S.
      • Charreire H.
      • Vergnaud A.C.
      • Hercberg S.
      • Oppert J.M.
      Changes in leisure-time physical activity and sedentary behaviour at retirement: a prospective study in middle-aged French subjects.
      More generally, associations between retirement and lifestyle behaviors tend to vary by sociodemographic characteristics and by pre-retirement lifestyle, job type, work stress, and voluntariness of retirement.
      • Kuerbis A.
      • Sacco P.
      The impact of retirement on the drinking patterns of older adults: a review.
      • Vahtera J.
      • Westerlund H.
      • Hall M.
      • et al.
      Effect of retirement on sleep disturbances: the GAZEL prospective cohort study.
      • Barnett I.
      • van Sluijs E.
      • Ogilvie D.
      • Wareham N.J.
      Changes in household, transport and recreational physical activity and television viewing time across the transition to retirement: longitudinal evidence from the EPIC-Norfolk cohort.
      • Henkens K.
      • van Solinge H.
      • Gallo W.T.
      Effects of retirement voluntariness on changes in smoking, drinking and physical activity among Dutch older workers.
      • Morris J.K.
      • Cook D.G.
      • Shaper A.G.
      Non-employment and changes in smoking, drinking, and body weight.
      • Richman J.A.
      • Zlatoper K.W.
      • Zackula Ehmke J.L.
      • Rospenda K.M.
      Retirement and drinking outcomes: lingering effects of workplace stress?.
      Using data from a large population-based Australian cohort, this study examined
      • 1
        the association between retirement and changes in health-related lifestyles behaviors; and
      • 2
        whether the changes differ by sociodemographic characteristics and reasons for retirement.
      Both traditional (smoking, alcohol use, physical activity, and diet) and emergent lifestyle risk behaviors (sedentary behavior, sleep) were assessed singularly and jointly.

      Methods

      Sampling and Procedures

      Participants were from the Social, Economic, and Environmental Factor study (SEEF), a follow-up of a subsample of the Sax Institute’s 45 and Up Study. The latter study comprised 267,153 adults aged ≥45 years from the state of New South Wales, Australia, surveyed between February 2006 and December 2009 (response rate, 18%).
      • Banks E.
      • Redman S.
      • Jorm L.
      • et al.
      Cohort profile: the 45 and Up Study.
      The first 100,000 respondents were invited to participate in SEEF in 2010 (response rate, 64.4%). All participants completed consent forms for both surveys. The baseline data collection and SEEF were approved by the University of New South Wales Human Research Ethics Committee (reference, HREC 05035) and the University of Sydney Human Research Ethics Committee (reference, 10–2009/12187).

      Measures

      Baseline questionnaires of the 45 and Up Study can be found at www.saxinstitute.org.au/our-work/45-up-study/questionnaires/. The SEEF questionnaire can be provided upon request.
      Dependent variables were six lifestyle behaviors measured at both baseline and follow-up. Responses were coded as 1 for being “at risk” and 0 for “not at risk” as follows:
      Current smoking (answering yes to both questions: Have you ever been a regular smoker? and Are you a regular smoker now?) was classified as “at risk.” Respondents reported the total number of alcoholic drinks they consumed each week (one drink defined as one glass of wine, half a pint of beer, or one shot of spirits). “At risk” drinking was defined as consuming >14 drinks/week, based on current Australian Guidelines.
      Australian Government National Health and Medical Research Council
      Australian Guidelines to Reduce Health Risks From Drinking Alcohol.
      Respondents were asked how many servings of fruit and vegetables they usually consumed each day. Not meeting Australian recommendations for fruit and vegetable consumption (i.e., two servings of fruit and five servings of vegetables) was used as a marker for dietary risk.
      Australian Government Department of Health
      Eat for Health: Australian Dietary Guidelines.
      The Active Australia Questionnaire assessed total time spent on walking and moderate- and vigorous-intensity physical activity (in bouts of ≥10 minutes) in the last week.
      Australian Institute of Health and Welfare (AIHW)
      The Active Australia Survey: A Guide and Manual for Implementation, Analysis and Reporting.
      This instrument has been found to have acceptable reliability and validity.
      • Brown W.J.
      • Burton N.W.
      • Marshall A.L.
      • Miller Y.D.
      Reliability and validity of a modified self-administered version of the Active Australia physical activity survey in a sample of mid-age women.
      Total moderate- to vigorous-intensity physical activity was calculated as the sum of the three types of activities with vigorous activity weighted by two.
      Australian Institute of Health and Welfare (AIHW)
      The Active Australia Survey: A Guide and Manual for Implementation, Analysis and Reporting.
      Reporting <150 minutes/week of moderate- to vigorous-intensity physical activity based on Australia’s Physical Activity and Sedentary Behaviour Guidelines
      Australian Governmet the Department of Health
      Australia’s Physical Activity and Sedentary Behaviour Guidelines.
      was considered as being “at risk” (insufficiently active).
      Sedentary behavior and sleep were measured by asking respondents how many hours in each 24-hour day they usually spent sitting and sleeping. More than 7 hours/day of sitting was considered “at risk” based on recent meta-analytic evidence.
      • Chau J.Y.
      • Grunseit A.C.
      • Chey T.
      • et al.
      Daily sitting time and all-cause mortality: a meta-analysis.
      Fewer than 7 or >9 hours/day of sleep was coded “at risk” (i.e., short/long sleep durations) based on meta-analyses on sleep and health outcomes.
      • Cappuccio F.P.
      • Cooper D.
      • D’Elia L.
      • Strazzullo P.
      • Miller M.A.
      Sleep duration predicts cardiovascular outcomes: a systematic review and meta-analysis of prospective studies.
      • Cappuccio F.P.
      • D’Elia L.
      • Strazzullo P.
      • Miller M.A.
      Sleep duration and all-cause mortality: a systematic review and meta-analysis of prospective studies.
      In addition to the single behavioral risk measures described above, a summary index was generated by summing the total number of risk behaviors for a score ranging from 0 to 6 to indicate respondents’ overall lifestyle risk.
      • Ding D.
      • Rogers K.
      • Macniven R.
      • et al.
      Revisiting lifestyle risk index assessment in a large Australian sample: should sedentary behavior and sleep be included as additional risk factors?.
      Respondents reported their working status at both baseline and follow-up by choosing one or more of the following options: in full-time paid work, in part-time paid work, self-employed, doing unpaid work, studying, completely retired/pensioner, partially retired, disabled/sick, looking after home/family, unemployed, and other. As respondents could select multiple options, a hierarchical coding approach was employed whereby some responses were prioritized over others to create mutually exclusive categories. Respondents’ retirement status was classified sequentially at baseline and follow-up. Respondents were classified as “not retired” if they
      • 1.
        chose in full-time paid work (irrespective of their responses to other categories); or
      • 2.
        selected in part-time paid work, but did not mark completely retired/pensioner; or
      • 3.
        selected unemployed or self-employed, but did not say they were completely retired or sick/disabled.
      Respondents were classified as “retired” if they
      • 1.
        selected completely retired/pensioner; or
      • 2.
        did not meet the definition above for “not retired.”
      Respondents who said they were retired were asked to indicate why they retired/stopped working. Similar to a previously published study,
      • Vo K.
      • Forder P.M.
      • Tavener M.
      • et al.
      Retirement, age, gender and mental health: findings from the 45 and Up Study.
      a sequential coding approach was used, prioritizing certain responses over others as follows:
      • 1.
        health problems (retired owing to own ill health or work-related injury);
      • 2.
        caring responsibilities (to care for sick/disabled person or to look after home/family);
      • 3.
        unemployment (made redundant, could not find a job, or was self-employed and business closed);
      • 4.
        age/lifestyle (reached usual retirement age, lifestyle reasons, to travel, or to study [combined to reflect voluntary reasons for retirement, as Australia does not have a mandatory retirement age]); and
      • 5.
        other.

      Statistical Analysis

      Analyses were restricted to those who were not retired (as defined above) at baseline.
      • Evenson K.R.
      • Rosamond WD
      • Cai J
      • et al.
      Influence of retirement on leisure-time physical activity: the Atherosclerosis Risk in Communities Study.
      Sample size at each stage of exclusion is presented in Appendix Figure 1 (available online). Final analytic sample sizes differed slightly because of missing values on lifestyle risk factors. The incident “retired” group (henceforth called “retirees”) included respondents classified as “not retired” at baseline and “retired” at follow-up. Respondents classified as “not retired” (henceforth called “non-retirees”) at both baseline and follow-up constituted the comparison group.
      Baseline sociodemographic characteristics and health behaviors of the two groups were compared using t-tests, chi-square tests, and age-adjusted logistic regression. Binary logistic regression models were fitted for each dichotomous risk behavior, adjusted for the baseline value of each outcome; duration of follow-up; age; sex; marital status; educational attainment; residential location (major city versus regional/remote, based on the Accessibility/Remoteness Index of Australia
      Australian Government Department of Health
      Measuring Remoteness. Accessibility/Remoteness Index of Australia (ARIA).
      ); and self-rated health (measured by a single-item health rating question in SF-12
      • Sanderson K.
      • Andrews G.
      The SF-12 in the Australian population: cross-validation of item selection.
      ). In addition to AORs for the association between retirement status and each risk behavior, estimated marginal probabilities are reported at follow-up by retirement status, indicating the predicted probability of having a risk factor at follow-up at the mean values of the covariates. Multiple linear regression models estimated change in lifestyle risk index score from baseline to follow-up with retirement status at follow-up as the independent variable, adjusted for baseline index score and other covariates. Estimated marginal means for these analyses represent the mean change in the index score at the mean values of the covariates according to the regression model. Because of the large proportion of participants being sufficiently active and consuming insufficient fruit and vegetables, further sensitivity analyses were conducted using 300 minutes/week of moderate- to vigorous-intensity physical activity
      Australian Governmet the Department of Health
      Australia’s Physical Activity and Sedentary Behaviour Guidelines.
      • Gebel K.
      • Ding D.
      • Chey T.
      • Stamatakis E.
      • Brown W.J.
      • Bauman A.E.
      Effect of moderate to vigorous physical activity on all-cause mortality in middle-aged and older Australians.
      and two servings of fruit plus three servings of vegetables
      • Ding D.
      • Do A.
      • Schmidt H.M.
      • Bauman A.E.
      A widening gap? Changes in multiple lifestyle risk behaviours by socioeconomic status in New South Wales, Australia, 2002-2012.
      as the at-risk cut-points. In addition, owing to the widely described J- or U-shaped relationship between alcohol and mortality,
      • Kloner R.A.
      • Rezkalla S.H.
      To drink or not to drink? That is the question.
      additional sensitivity analysis was conducted coding those who consumed zero servings of alcohol to be “at risk.” Finally, the authors repeated the analysis using a mixed model (repeated measure) approach, which may be less biased in non-randomized studies.
      • Van Breukelen G.J.
      ANCOVA versus change from baseline: more power in randomized studies, more bias in nonrandomized studies [corrected].
      Sociodemographic characteristics, including age (45–64 years versus ≥65 years at baseline); sex; area of residence (major city versus regional/remote); educational attainment; and country of birth (Australia born versus foreign born) were tested as potential effect modifiers as suggested by the literature.
      • Touvier M.
      • Bertrais S.
      • Charreire H.
      • Vergnaud A.C.
      • Hercberg S.
      • Oppert J.M.
      Changes in leisure-time physical activity and sedentary behaviour at retirement: a prospective study in middle-aged French subjects.
      • Barnett I.
      • Guell C.
      • Ogilvie D.
      How do couples influence each other’s physical activity behaviours in retirement? An exploratory qualitative study.
      In addition, total hours of work per week at baseline was tested as an additional potential effect modifier. Hours worked were categorized into <35 hours versus ≥35 hours/week to correspond with full-time/part-time status in the Australian labor force.

      Australian Bureau of Statistics. Understanding full-time/part-time status in the labour force survey. www.abs.gov.au/ausstats/[email protected]/Previousproducts/6202.0Main%20Features4Sep%202013?opendocument&tabname=Summary&prodno=6202.0&issue=Sep%202013&num=&view. Published 2014. Accessed October 30, 2015.

      Finally, additional regression models were fitted for incident retirees only with reason for retirement as the independent variable, adjusted for all covariates. All statistical analyses were conducted using Stata, version 13, and significance levels were set at p<0.05.

      Results

      During the 3.3 (SD=0.9) years of follow-up, of the 27,257 working at baseline, 3,106 respondents retired from full-/part-time work (11.4%; Table 1). Retirees were significantly older, less likely to have a university degree, and more likely to have fair/poor health at baseline.
      Table 1Baseline Demographic and Lifestyle Characteristics by Retirement Status at Follow-up
      CharacteristicWorking (n=24,151)Retired (n=3,106)p-value
      Comparison between those who were working and retired based on independent sample t-test for the continuous variable (age), and χ2 test for categorical variables.
      p-value
      Comparison between those who were working and retired based on logistic regression adjusted for age. HSC, Higher School Certificate; TAFE, Technical and Further Education.
      Demographic characteristics (%)
       Age, M (SD) year54.3 (6.2)62.4 (6.7)<0.001
       Female50.749.3<0.001<0.001
       Educational attainment<0.001<0.001
        Up to 10 years20.929.8
        HSC/TAFE/diploma43.742.9
        Degree35.527.2
       Married/de facto81.780.30.0730.022
       Living in major cities45.044.30.4840.620
       Fair/poor health7.710.6<0.001<0.001
      Health risk behavior at baseline (%)
       Smoking6.75.70.0240.153
       Excessive alcohol use15.717.30.0200.868
       Insufficient fruit and vegetable consumption80.076.4<0.001<0.001
       Insufficient physical activity18.419.40.1500.100
       Short/long sleep durations17.621.5<0.0010.004
       Excessive sitting31.927.7<0.0010.488
      Health risk behavior at follow-up (%)
       Smoking5.63.7<0.0010.317
       Excessive alcohol use15.616.80.090.480
       Insufficient fruit and vegetable consumption79.275.3<0.001<0.001
       Insufficient physical activity16.914.60.002<0.001
       Short/long sleep durations19.420.80.080.181
       Excessive sitting27.810.6<0.001<0.001
      Note: Boldface indicates statistical significance (p<0.05).
      a Comparison between those who were working and retired based on independent sample t-test for the continuous variable (age), and χ2 test for categorical variables.
      b Comparison between those who were working and retired based on logistic regression adjusted for age.HSC, Higher School Certificate; TAFE, Technical and Further Education.
      As shown in Table 2, adjusted for covariates, retirees had significantly lower odds for a range of risk factors at follow-up, including insufficient physical activity (AOR=0.73, p<0.001); excessive sitting (AOR=0.34, p<0.001); and short/long sleep durations (AOR=0.82, p<0.001). The association of retirement with smoking was marginal (AOR=0.74, p<0.05). There was no significant association between retirement status and alcohol use or fruit and vegetable consumption at follow-up. Retirees had a larger reduction in the overall lifestyle risk index score compared with non-retirees (–0.27 vs –0.07, p<0.001). Supplementary analyses using a mixed model with a time by retirement status interaction term
      • Van Breukelen G.J.
      ANCOVA versus change from baseline: more power in randomized studies, more bias in nonrandomized studies [corrected].
      showed very similar effect sizes and significance.
      Table 2AORs
      Based on binary logistic regression adjusted for follow-up time, age, sex, educational attainment, marital status, general self-rated health, area of residence (major city/regional/remote), and the baseline measure of each outcome behavior. Interpreted as the probability of having a risk factor at follow-up at the mean values of covariates for those who had retired and those who were still working.
      and Marginal Probability of Risk Behaviors by Retirement Status (2006–2010)
      RetiredWorking
      Risk behaviorRetired versus not retired, OR (95% CI)nMarginal probability
      Based on binary logistic regression adjusted for follow-up time, age, sex, educational attainment, marital status, general self-rated health, area of residence (major city/regional/remote), and the baseline measure of each outcome behavior. Interpreted as the probability of having a risk factor at follow-up at the mean values of covariates for those who had retired and those who were still working.
      nMarginal probability
      Based on binary logistic regression adjusted for follow-up time, age, sex, educational attainment, marital status, general self-rated health, area of residence (major city/regional/remote), and the baseline measure of each outcome behavior. Interpreted as the probability of having a risk factor at follow-up at the mean values of covariates for those who had retired and those who were still working.
      Smoking0.74 (0.55, 0.99)*3,0490.05 (0.04, 0.05)23,8460.06 (0.05, 0.06)
      Excessive alcohol use1.17 (1.00, 1.36)2,9620.17 (0.16, 0.18)23,2890.16 (0.15, 0.16)
      Insufficient fruit and vegetable consumption0.92 (0.83, 1.02)3,0490.78 (0.76, 0.79)23,8460.79 (0.78, 0.80)
      Insufficient physical activity0.73 (0.65, 0.83)***2,9360.13 (0.12, 0.14)22,9960.17 (0.16, 0.17)
      Excessive sitting0.34 (0.29, 0.39)***2,7410.13 (0.12, 0.15)22,1860.27 (0.27, 0.28)
      Short/long sleep durations0.82 (0.72, 0.91)***2,9490.17 (0.16, 0.19)23,3200.20 (0.19, 0.20)
      B (95% CI)nMarginal mean
      Based on general linear regression adjusted for follow-up time, age, sex, educational attainment, marital status, general self-rated health, area of residence (major city/regional/remote), and the baseline measure of the lifestyle risk index. Interpreted as the mean change in lifestyle risk index from baseline to follow-up at the mean values of covariates for those who had retired and those who were still working.
      nMarginal mean
      Based on general linear regression adjusted for follow-up time, age, sex, educational attainment, marital status, general self-rated health, area of residence (major city/regional/remote), and the baseline measure of the lifestyle risk index. Interpreted as the mean change in lifestyle risk index from baseline to follow-up at the mean values of covariates for those who had retired and those who were still working.
      Change in lifestyle risk index–0.24 (–0.28, –0.20)***2,569–0.27 (–0.31, –0.24)20,909–0.07 (–0.08, –0.06)
      Note: Boldface indicates statistical significance (*p<0.05; ***p<0.001).
      a Based on binary logistic regression adjusted for follow-up time, age, sex, educational attainment, marital status, general self-rated health, area of residence (major city/regional/remote), and the baseline measure of each outcome behavior. Interpreted as the probability of having a risk factor at follow-up at the mean values of covariates for those who had retired and those who were still working.
      b Based on general linear regression adjusted for follow-up time, age, sex, educational attainment, marital status, general self-rated health, area of residence (major city/regional/remote), and the baseline measure of the lifestyle risk index. Interpreted as the mean change in lifestyle risk index from baseline to follow-up at the mean values of covariates for those who had retired and those who were still working.
      Additional linear regression models were fitted to quantify absolute change in time use behaviors (i.e., physical activity, sitting, and sleep time) from baseline to follow-up. Compared with non-retirees and adjusted for all covariates, retirees reported a significantly larger increase in time spent walking (33 vs 16 minutes/week, p<0.001) and in moderate-intensity physical activity (59 vs 24 minutes/week, p<0.001) during the follow-up period. The change in vigorous-intensity physical activity did not differ by retirement status (1 vs –4 minutes/week, p=0.110). Retirees also reported a larger decrease in sitting time (–67 vs –27 minutes/day, p<0.001) and a larger increase in sleep duration (11.1 vs –4.2 min/day, p<0.001), compared with non-retirees.
      Several interactions (effect modifications) were significant (Table 3). Overall, the protective effect of retirement tended to be stronger among the middle-aged than older participants, and those who worked full-time prior to retirement. Retirement was protective against smoking only in women, not in men, and the inverse association between retirement and excessive sitting was stronger among those who lived in major cities and as education increased.
      Table 3Significant Effect Modifiers for the Association Between Retirement and Lifestyle Changes (2006–2010)
      p-value for interaction, or OR (95% CI) for retired versus not retired
      VariableSmokingInsufficient physical activityExcessive sittingShort/long sleep durationLifestyle risk index, p-value for interaction, or B (95% CI)
      Age (years)p<0.001p<0.001p<0.001
       45–640.67 (0.60, 0.75)0.29 (0.24, 0.34)–0.27 (–0.31, –0.23)
       ≥650.96 (0.77, 1.18)0.67 (0.50, 0.90)–0.08 (–0.17, 0.01)
      Sexp=0.013
       Female0.53 (0.34, 0.82)
       Male0.99 (0.67, 1.47)
      Educationp=0.020
       Up to 10 years0.46 (0.35, 0.61)
       HSC/TAFE/diploma0.33 (0.27, 0.42)
       Degree0.28 (0.22, 0.35)
      Residential locationp=0.001
       Major city0.26 (0.22, 0.32)
       Regional/remote0.44 (0.36, 0.53)
      Pre-retirement working time (hours/week)p<0.001p<0.001p<0.001p<0.001
       <350.86 (0.75, 0.99)0.60 (0.49, 0.73)1.03 (0.87, 1.21)–0.09 (–0.15, –0.03)
       ≥350.65 (0.54, 0.79)0.20 (0.16, 0.24)0.58 (0.48, 0.70)–0.42 (–0.48, –0.36)
      Note: Boldface indicates statistical significance (p<0.05). Blank cells indicate non-significant effect modifications/interactions.
      Country of birth was tested as an effect modifier but it was not significant in any model; therefore, it was deleted from Table 3.
      All potential effect modifiers were tested for the association of retirement with alcohol use and fruit and vegetable assumption. None of the tested effect modifiers was significant, therefore these two outcomes were deleted from Table 3.
      HSC, Higher School Certificate; TAFE, Technical and Further Education.
      Among retirees (n=3,106), 47.2% retired because of age or lifestyle reasons, 20.5% because of lack of job opportunities, 17.4% owing to health problems, and 5.6% retired to take care of a family member or a friend. A further 7.8% retired for other reasons not further specified, and 1.5% did not provide any reason (missing). As shown in Table 4, compared with those who retired for age or for lifestyle reasons, those who retired because of caring responsibilities were more likely to be current smokers (AOR=3.45) and those who retired for health reasons were more likely to have short/long sleep durations and to sit excessively at follow-up (AOR=1.40 and 1.56, respectively). The change in the lifestyle risk index score also differed significantly by reason for retirement. Although there were significant reductions in the combined risk score for all subgroups, those who retired for health reasons experienced the smallest reduction.
      Table 4Change in Lifestyle Behaviors by Reason of Retirement Among Participants Who Retired During the Study
      AOR (95% CI)
      Based on binary logistic regression adjusted for follow-up time, age, sex, educational attainment, marital status, general self-rated health, area of residence (major city/regional/remote), and the baseline measure of the outcome behavior.
      Risk behaviorReached retirement age/lifestyle (n=1,467)Ill health/work injuries (n=541)Made redundant/could not find job/business closed(n=637)To care for family/friend(n=173)Other(n=241)p-value
      Smoking1.00 (ref)1.27 (0.59, 2.73)1.33 (0.66, 2.66)3.45 (1.23, 9.66)*0.63 (0.17, 2.37)0.155
      Excessive alcohol use1.00 (ref)1.05 (0.71, 1.57)1.17 (0.83, 1.64)0.67 (0.34, 1.32)1.00 (0.58, 1.75)0.633
      Insufficient fruit and vegetable consumption1.00 (ref)0.90 (0.69, 1.17)1.03 (0.81, 1.32)0.79 (0.54, 1.15)0.90 (0.64, 1.26)0.646
      Insufficient physical activity1.00 (ref)1.27 (0.94, 1.72)1.01 (0.75, 1.36)1.03 (0.62, 1.72)0.99 (0.64, 1.54)0.617
      Short/long sleep durations1.00 (ref)1.40 (1.06, 1.85)*0.98 (0.75, 1.27)1.13 (0.73, 1.76)0.92 (0.63, 1.35)0.128
      Excessive sitting1.00 (ref)1.56 (1.11, 2.21)*1.02 (0.72, 1.44)1.06 (0.58, 1.93)1.17 (0.71, 1.91)0.132
      Marginal mean change (95% CI)
      Based on general linear regression adjusted for follow-up time, age, sex, educational attainment, marital status, general self-rated health, area of residence (major city/regional/remote), and the baseline measure of the lifestyle risk index score.
      Lifestyle risk index–0.34 (–0.38, –0.29)–0.13 (–0.21, –0.05)***–0.28 (–0.35, –0.21)–0.28 (–0.42, –0.15)–0.40 (–0.52, –0.29)<0.001
      Note: Boldface indicates statistical significance (*p<0.05; ***p<0.001) as compared with those who retired at retirement age and/or for lifestyle reasons.
      a Based on binary logistic regression adjusted for follow-up time, age, sex, educational attainment, marital status, general self-rated health, area of residence (major city/regional/remote), and the baseline measure of the outcome behavior.
      b Based on general linear regression adjusted for follow-up time, age, sex, educational attainment, marital status, general self-rated health, area of residence (major city/regional/remote), and the baseline measure of the lifestyle risk index score.

      Discussion

      This large population-based Australian study found that transitioning to retirement was associated with improved profiles of health-related behaviors, characterized by less smoking (among women only), more physical activity, less sedentary behavior, and healthier sleep patterns. As one of the first studies to comprehensively examine a broad range of lifestyle behaviors in the context of retirement, the current study implies that retirement could represent a positive transition to a healthier lifestyle.
      In the present study, retirees reported significantly greater increase in physical activity, particularly in walking and moderate-intensity activities, perhaps reflecting the greater opportunity for active living that retirement can offer. Analyses further showed that this “activity-promoting effect” of retirement is likely to benefit those who retired at a younger age, possibly because of better physical function, and those who worked full-time prior to retirement, perhaps owing to a larger amount of time “gained” through retirement. However, one cannot infer that retirees became more active overall, because the Active Australia Questionnaire does not quantify occupational activity.
      Australian Institute of Health and Welfare (AIHW)
      The Active Australia Survey: A Guide and Manual for Implementation, Analysis and Reporting.
      Although previous studies
      • Touvier M.
      • Bertrais S.
      • Charreire H.
      • Vergnaud A.C.
      • Hercberg S.
      • Oppert J.M.
      Changes in leisure-time physical activity and sedentary behaviour at retirement: a prospective study in middle-aged French subjects.
      • Berger U.
      • Der G.
      • Mutrie N.
      • Hannah M.K.
      The impact of retirement on physical activity.
      • Brown W.J.
      • Heesch K.C.
      • Miller Y.D.
      Life events and changing physical activity patterns in women at different life stages.
      have found a substantial increase in leisure-time physical activity post-retirement, other research shows this does not always compensate for the decrease in occupational physical activity, and retirement is consequently associated with an actual decline in overall physical activity.
      • Barnett I.
      • van Sluijs E.
      • Ogilvie D.
      • Wareham N.J.
      Changes in household, transport and recreational physical activity and television viewing time across the transition to retirement: longitudinal evidence from the EPIC-Norfolk cohort.
      • Berger U.
      • Der G.
      • Mutrie N.
      • Hannah M.K.
      The impact of retirement on physical activity.
      To date, very few studies have examined the impact of retirement on sedentary behavior,
      • Touvier M.
      • Bertrais S.
      • Charreire H.
      • Vergnaud A.C.
      • Hercberg S.
      • Oppert J.M.
      Changes in leisure-time physical activity and sedentary behaviour at retirement: a prospective study in middle-aged French subjects.
      • Barnett I.
      • van Sluijs E.
      • Ogilvie D.
      • Wareham N.J.
      Changes in household, transport and recreational physical activity and television viewing time across the transition to retirement: longitudinal evidence from the EPIC-Norfolk cohort.
      • Menai M.
      • Fezeu L.
      • Charreire H.
      • et al.
      Changes in sedentary behaviours and associations with physical activity through retirement: a 6-year longitudinal study.
      an emergent lifestyle risk factor.
      • Bauman A.
      • Chau J.
      • Ding D.
      • Bennie J.
      Too much sitting and cardio-metabolic risk: an update of epidemiological evidence.
      Three European studies found that retirement was associated with increased TV viewing time
      • Touvier M.
      • Bertrais S.
      • Charreire H.
      • Vergnaud A.C.
      • Hercberg S.
      • Oppert J.M.
      Changes in leisure-time physical activity and sedentary behaviour at retirement: a prospective study in middle-aged French subjects.
      • Barnett I.
      • van Sluijs E.
      • Ogilvie D.
      • Wareham N.J.
      Changes in household, transport and recreational physical activity and television viewing time across the transition to retirement: longitudinal evidence from the EPIC-Norfolk cohort.
      • Menai M.
      • Fezeu L.
      • Charreire H.
      • et al.
      Changes in sedentary behaviours and associations with physical activity through retirement: a 6-year longitudinal study.
      and overall leisure-time sedentary behavior,
      • Menai M.
      • Fezeu L.
      • Charreire H.
      • et al.
      Changes in sedentary behaviours and associations with physical activity through retirement: a 6-year longitudinal study.
      possibly explained by increased disposable time at home. By contrast, the current study found that retirement was associated with a substantial decrease in overall sitting time, with the largest reductions occurring among younger retirees, those living in major cities, those with higher levels of education, and who worked full-time prior to retirement. Research suggests that prolonged sitting is associated with higher education levels
      • Bauman A.
      • Ainsworth B.E.
      • Sallis J.F.
      • et al.
      The descriptive epidemiology of sitting. A 20-country comparison using the International Physical Activity Questionnaire (IPAQ).
      and type of occupations and job role.
      • Chau J.Y.
      • van der Ploeg H.P.
      • Merom D.
      • Chey T.
      • Bauman A.E.
      Cross-sectional associations between occupational and leisure-time sitting, physical activity and obesity in working adults.
      • Jans M.P.
      • Proper K.I.
      • Hildebrandt V.H.
      Sedentary behavior in Dutch workers: differences between occupations and business sectors.
      Thus, retirement eliminates sitting accumulated at work (particularly among full-time employees) and during the commute to and from work.
      • Jans M.P.
      • Proper K.I.
      • Hildebrandt V.H.
      Sedentary behavior in Dutch workers: differences between occupations and business sectors.
      • Sugiyama T.
      • Ding D.
      • Owen N.
      Commuting by car: weight gain among physically active adults.
      One may speculate that workers retiring from high sitting jobs may be more likely to “benefit” from retirement in terms of reduced total sitting time.
      The current study also found that retirement was associated with reduced odds of smoking in women, but not in men. Similarly, a recent study based on the French GAZEL cohort found that women, but not men, had decreased odds of smoking after 5 years as compared with 1 year since retirement.
      • Tamers S.L.
      • Okechukwu C.
      • Marino M.
      • Guéguen A.
      • Goldberg M.
      • Zins M.
      Effect of stressful life events on changes in smoking among the French: longitudinal findings from GAZEL.
      Based on the literature, mixed evidence exists on whether women are more likely to quit smoking in response to life events
      • Tamers S.L.
      • Okechukwu C.
      • Marino M.
      • Guéguen A.
      • Goldberg M.
      • Zins M.
      Effect of stressful life events on changes in smoking among the French: longitudinal findings from GAZEL.
      • McKee S.A.
      • Maciejewski P.K.
      • Falba T.
      • Mazure C.M.
      Sex differences in the effects of stressful life events on changes in smoking status.
      ; hence, it is difficult to conclude whether retirement has beneficial effects on smoking.
      It was found that retirees increased their sleep time by 15 minutes/day compared with non-retirees, and were more likely to get healthy amounts of sleep. This finding echoes previous research on retirement and sleep outcomes.
      • Vahtera J.
      • Westerlund H.
      • Hall M.
      • et al.
      Effect of retirement on sleep disturbances: the GAZEL prospective cohort study.
      • Mehra R.
      Retire for better sleep?.
      Together, these studies suggest that retirement is likely to lead to healthier sleep patterns.
      Finally, alcohol use and fruit and vegetable consumption were not associated with retirement in the current study, and both findings remained robust after sensitivity analyses. A recent review on retirement and alcohol use found a lack of association overall,
      • Kuerbis A.
      • Sacco P.
      The impact of retirement on the drinking patterns of older adults: a review.
      and the literature on retirement and diet is too sparse to make effective comparisons with the current study.
      • Zantinge E.M.
      • van den Berg M.
      • Smit H.A.
      • Picavet HSJ
      Retirement and a healthy lifestyle: opportunity or pitfall? A narrative review of the literature.
      The observed post-retirement lifestyle changes are not unexpected. As job-related activities consume a substantial proportion of people’s time, and job strain is a risk factor for coronary heart disease,
      • Kivimäki M.
      • Nyberg S.T.
      • Batty G.D.
      • et al.
      Job strain as a risk factor for coronary heart disease: a collaborative meta-analysis of individual participant data.
      • Kivimäki M.
      • Virtanen M.
      • Elovainio M.
      • Kouvonen A.
      • Väänänen A.
      • Vahtera J.
      Work stress in the etiology of coronary heart disease—a meta-analysis.
      retirement removes these factors and therefore is likely to be health and well-being enhancing. It is important to note that the association between retirement and subsequent lifestyle changes is likely to be context specific, and may particularly depend on life circumstances and the type of job held prior to retirement. In this study, a stronger association was found between retirement and positive lifestyle changes among those who worked full-time prior to retirement. This may be explained by the lack of time particularly among those who were working full-time.
      The literature suggests that health and lifestyle benefits of retirement may differ by reason for retirement.
      • Henkens K.
      • van Solinge H.
      • Gallo W.T.
      Effects of retirement voluntariness on changes in smoking, drinking and physical activity among Dutch older workers.
      • Barnett I.
      • van Sluijs E.M.
      • Ogilvie D.
      Physical activity and transitioning to retirement: a systematic review.
      Overall, retirement was associated with reductions in lifestyle risk behaviors across different reasons for retirement, but the reduction was significantly attenuated among those who retired because of illness or injury. Specifically, compared with those who retired for age or lifestyle reasons, those who retired because of illness/injury were less likely to be sufficiently active and to have healthy sleep duration at follow-up, possibly because of the deterioration of health. Those who retired to care for a family/friend had more than three times the odds of smoking as compared with those who retired voluntarily, perhaps highlighting the chronic stress associated with caregiving.

      Limitations

      The current study analyzed prospective data from a large population sample. The novelty lies in examining a broad range of lifestyle behaviors both in terms of individual and combined health risk. This study also examined sociodemographic characteristics, working hours prior to retirement, and reasons for retirement as potential effect modifiers, and the robustness of findings is enhanced through several sensitivity analyses. Interpretation of these findings is limited by lack of information on type of pre-retirement occupation, as previous studies
      • Touvier M.
      • Bertrais S.
      • Charreire H.
      • Vergnaud A.C.
      • Hercberg S.
      • Oppert J.M.
      Changes in leisure-time physical activity and sedentary behaviour at retirement: a prospective study in middle-aged French subjects.
      • Chung S.
      • Domino ME
      • Stearns SC
      • Popkin BM
      Retirement and physical activity: analyses by occupation and wealth.
      have found that change in lifestyle behaviors depends on the nature of the pre-retirement job. Domain-specific measures, such as occupational physical activity and TV viewing, may have also assisted in elucidating underlying reasons for the observed changes. Given that the follow-up time was relatively short, a longer follow-up period would help track whether the observed changes are maintained, as a previous study
      • Jokela M.
      • Ferrie J.E.
      • Gimeno D.
      • et al.
      From midlife to early old age: Health trajectories associated with retirement.
      found that the health improvements after voluntary retirement attenuated over time. Although the 45 and Up Study had a low response rate, a comparison of this cohort with a representative sample of the New South Wales general population found similar estimates for risk factor–outcome relationships.
      • Mealing N.
      • Banks E.
      • Jorm L.
      • Steel D.
      • Clements M.
      • Rogers K.
      Investigation of relative risk estimates from studies of the same population with contrasting response rates and designs.
      Finally, it is important to note that the association between retirement and health-related lifestyles found in the current study may not be generalizable to other countries with different retirement systems, health insurance schemes, and social safety nets.

      Conclusions

      In a large population-based Australian cohort, retirement was associated with positive lifestyle changes, at least in the short term. Programs for retirees may capitalize on the healthy transitions through retirement.

      Acknowledgments

      DD is funded by an Early Career Fellowship from the Australian National Health and Medical Research Council (#1072223). JYC is supported by a Postdoctoral Fellowship (#100567) from the National Heart Foundation of Australia. AB was funded by a Program Grant (#569940) from the Australian National Health and Medical Research Council. The funder did not play any role in the study design, data collection, analysis and interpretation of data, writing, or the decision to submit the paper for publication.
      No financial disclosures were reported by the authors of this paper.

      Appendix A. Supplementary materials

      References

        • WHO
        10 Facts on Ageing and the Life Course.
        WHO, Geneva, Switzerland2014
        • Prince M.J.
        • Wu F.
        • Guo Y.
        • et al.
        The burden of disease in older people and implications for health policy and practice.
        Lancet. 2015; 385: 549-562https://doi.org/10.1016/S0140-6736(14)61347-7
        • Knoops K.T.
        • de Groot LC
        • Kromhout D
        • et al.
        Mediterranean diet, lifestyle factors, and 10-year mortality in elderly European men and women: the HALE project.
        JAMA. 2004; 292: 1433-1439https://doi.org/10.1001/jama.292.12.1433
        • King D.E.
        • Mainous 3rd, A.G.
        • Geesey M.E.
        Turning back the clock: adopting a healthy lifestyle in middle age.
        Am J Med. 2007; 120: 598-603https://doi.org/10.1016/j.amjmed.2006.09.020
        • Kim J.E.
        • Moen P.
        Retirement transitions, gender, and psychological well-being: a life-course, ecological model.
        J Gerontol B Psychol Sci Soc Sci. 2002; 57: P212-P222https://doi.org/10.1093/geronb/57.3.P212
        • Wang J.
        • Lin W.
        • Chang L.
        The relationships between retirement, social capital, and self-perceived health.
        Gerontologist. 2010; 50: 346-347
        • Jonsson H.
        • Josephsson S.
        • Kielhofner G.
        Narratives and experience in an occupational transition: a longitudinal study of the retirement process.
        Am J Occup Ther. 2001; 55: 424-432https://doi.org/10.5014/ajot.55.4.424
        • Barnett I.
        • Guell C.
        • Ogilvie D.
        The experience of physical activity and the transition to retirement: a systematic review and integrative synthesis of qualitative and quantitative evidence.
        Int J Behav Nutr Phys Act. 2012; 9: 97https://doi.org/10.1186/1479-5868-9-97
        • Kuerbis A.
        • Sacco P.
        The impact of retirement on the drinking patterns of older adults: a review.
        Addict Behav. 2012; 37: 587-595https://doi.org/10.1016/j.addbeh.2012.01.022
        • Zantinge E.M.
        • van den Berg M.
        • Smit H.A.
        • Picavet HSJ
        Retirement and a healthy lifestyle: opportunity or pitfall? A narrative review of the literature.
        Eur J Public Health. 2014; 24: 433-439https://doi.org/10.1093/eurpub/ckt157
        • Vahtera J.
        • Westerlund H.
        • Hall M.
        • et al.
        Effect of retirement on sleep disturbances: the GAZEL prospective cohort study.
        Sleep. 2009; 32: 1459-1466
        • Marquiae J.C.
        • Folkard S.
        • Ansiau D.
        • Tucker P.
        Effects of age, gender, and retirement on perceived sleep problems: results from the VISAT combined longitudinal and cross-sectional study.
        Sleep. 2012; 35: 1115-1121https://doi.org/10.5665/sleep.2000
        • Touvier M.
        • Bertrais S.
        • Charreire H.
        • Vergnaud A.C.
        • Hercberg S.
        • Oppert J.M.
        Changes in leisure-time physical activity and sedentary behaviour at retirement: a prospective study in middle-aged French subjects.
        Int J Behav Nutr Phys Act. 2010; 7: 14https://doi.org/10.1186/1479-5868-7-14
        • Barnett I.
        • van Sluijs E.
        • Ogilvie D.
        • Wareham N.J.
        Changes in household, transport and recreational physical activity and television viewing time across the transition to retirement: longitudinal evidence from the EPIC-Norfolk cohort.
        J Epidemiol Community Health. 2014; 68: 747-753https://doi.org/10.1136/jech-2013-203225
        • Henkens K.
        • van Solinge H.
        • Gallo W.T.
        Effects of retirement voluntariness on changes in smoking, drinking and physical activity among Dutch older workers.
        Eur J Public Health. 2008; 18: 644-649https://doi.org/10.1093/eurpub/ckn095
        • Morris J.K.
        • Cook D.G.
        • Shaper A.G.
        Non-employment and changes in smoking, drinking, and body weight.
        BMJ. 1992; 304: 536-541https://doi.org/10.1136/bmj.304.6826.536
        • Richman J.A.
        • Zlatoper K.W.
        • Zackula Ehmke J.L.
        • Rospenda K.M.
        Retirement and drinking outcomes: lingering effects of workplace stress?.
        Addict Behav. 2006; 31: 767-776https://doi.org/10.1016/j.addbeh.2005.06.001
        • Banks E.
        • Redman S.
        • Jorm L.
        • et al.
        Cohort profile: the 45 and Up Study.
        Int J Epidemiol. 2008; 37: 941-947https://doi.org/10.1093/ije/dym184
        • Australian Government National Health and Medical Research Council
        Australian Guidelines to Reduce Health Risks From Drinking Alcohol.
        Australian Government National Health and Medical Research Council, Canberra ACT, Australia2009
        • Australian Government Department of Health
        Eat for Health: Australian Dietary Guidelines.
        Australian Government Department of Health, Canberra ACT, Australia2013
        • Australian Institute of Health and Welfare (AIHW)
        The Active Australia Survey: A Guide and Manual for Implementation, Analysis and Reporting.
        AIHW, Canberra ACT, Australia2003
        • Brown W.J.
        • Burton N.W.
        • Marshall A.L.
        • Miller Y.D.
        Reliability and validity of a modified self-administered version of the Active Australia physical activity survey in a sample of mid-age women.
        Aust N Z J Public Health. 2008; 32: 535-541https://doi.org/10.1111/j.1753-6405.2008.00305.x
        • Australian Governmet the Department of Health
        Australia’s Physical Activity and Sedentary Behaviour Guidelines.
        The Department of Health, Canberra ACT, Australia2014
        • Chau J.Y.
        • Grunseit A.C.
        • Chey T.
        • et al.
        Daily sitting time and all-cause mortality: a meta-analysis.
        PLoS One. 2013; 8: e80000https://doi.org/10.1371/journal.pone.0080000
        • Cappuccio F.P.
        • Cooper D.
        • D’Elia L.
        • Strazzullo P.
        • Miller M.A.
        Sleep duration predicts cardiovascular outcomes: a systematic review and meta-analysis of prospective studies.
        Eur Heart J. 2011; 32: 1484-1492https://doi.org/10.1093/eurheartj/ehr007
        • Cappuccio F.P.
        • D’Elia L.
        • Strazzullo P.
        • Miller M.A.
        Sleep duration and all-cause mortality: a systematic review and meta-analysis of prospective studies.
        Sleep. 2010; 33: 585-592
        • Ding D.
        • Rogers K.
        • Macniven R.
        • et al.
        Revisiting lifestyle risk index assessment in a large Australian sample: should sedentary behavior and sleep be included as additional risk factors?.
        Prev Med. 2014; 60: 102-106https://doi.org/10.1016/j.ypmed.2013.12.021
        • Vo K.
        • Forder P.M.
        • Tavener M.
        • et al.
        Retirement, age, gender and mental health: findings from the 45 and Up Study.
        Aging Ment Health. 2015; 19: 647-657https://doi.org/10.1080/13607863.2014.962002
        • Evenson K.R.
        • Rosamond WD
        • Cai J
        • et al.
        Influence of retirement on leisure-time physical activity: the Atherosclerosis Risk in Communities Study.
        Am J Epidemiol. 2002; 155: 692-699https://doi.org/10.1093/aje/155.8.692
        • Australian Government Department of Health
        Measuring Remoteness. Accessibility/Remoteness Index of Australia (ARIA).
        Australian Government Department of Health, Canberra ACT, Australia2001
        • Sanderson K.
        • Andrews G.
        The SF-12 in the Australian population: cross-validation of item selection.
        Aust N Z J Public Health. 2002; 26: 343-345https://doi.org/10.1111/j.1467-842X.2002.tb00182.x
        • Gebel K.
        • Ding D.
        • Chey T.
        • Stamatakis E.
        • Brown W.J.
        • Bauman A.E.
        Effect of moderate to vigorous physical activity on all-cause mortality in middle-aged and older Australians.
        JAMA Intern Med. 2015; 175: 970-977https://doi.org/10.1001/jamainternmed.2015.0541
        • Ding D.
        • Do A.
        • Schmidt H.M.
        • Bauman A.E.
        A widening gap? Changes in multiple lifestyle risk behaviours by socioeconomic status in New South Wales, Australia, 2002-2012.
        PLoS One. 2015; 10: e0135338https://doi.org/10.1371/journal.pone.0135338
        • Kloner R.A.
        • Rezkalla S.H.
        To drink or not to drink? That is the question.
        Circulation. 2007; 116: 1306-1317https://doi.org/10.1161/CIRCULATIONAHA.106.678375
        • Van Breukelen G.J.
        ANCOVA versus change from baseline: more power in randomized studies, more bias in nonrandomized studies [corrected].
        J Clin Epidemiol. 2006; 59: 920-925https://doi.org/10.1016/j.jclinepi.2006.02.007
        • Barnett I.
        • Guell C.
        • Ogilvie D.
        How do couples influence each other’s physical activity behaviours in retirement? An exploratory qualitative study.
        BMC Public Health. 2013; 13: 1197https://doi.org/10.1186/1471-2458-13-1197
      1. Australian Bureau of Statistics. Understanding full-time/part-time status in the labour force survey. www.abs.gov.au/ausstats/[email protected]/Previousproducts/6202.0Main%20Features4Sep%202013?opendocument&tabname=Summary&prodno=6202.0&issue=Sep%202013&num=&view. Published 2014. Accessed October 30, 2015.

        • Berger U.
        • Der G.
        • Mutrie N.
        • Hannah M.K.
        The impact of retirement on physical activity.
        Ageing Soc. 2005; 25: 181-195https://doi.org/10.1017/S0144686X04002739
        • Brown W.J.
        • Heesch K.C.
        • Miller Y.D.
        Life events and changing physical activity patterns in women at different life stages.
        Ann Behav Med. 2009; 37: 294-305https://doi.org/10.1007/s12160-009-9099-2
        • Menai M.
        • Fezeu L.
        • Charreire H.
        • et al.
        Changes in sedentary behaviours and associations with physical activity through retirement: a 6-year longitudinal study.
        PLoS One. 2014; 9: e106850https://doi.org/10.1371/journal.pone.0106850
        • Bauman A.
        • Chau J.
        • Ding D.
        • Bennie J.
        Too much sitting and cardio-metabolic risk: an update of epidemiological evidence.
        Curr Cardiovasc Risk Rep. 2013; 7: 293-298https://doi.org/10.1007/s12170-013-0316-y
        • Bauman A.
        • Ainsworth B.E.
        • Sallis J.F.
        • et al.
        The descriptive epidemiology of sitting. A 20-country comparison using the International Physical Activity Questionnaire (IPAQ).
        Am J Prev Med. 2011; 41: 228-235https://doi.org/10.1016/j.amepre.2011.05.003
        • Chau J.Y.
        • van der Ploeg H.P.
        • Merom D.
        • Chey T.
        • Bauman A.E.
        Cross-sectional associations between occupational and leisure-time sitting, physical activity and obesity in working adults.
        Prev Med. 2012; 54: 195-200https://doi.org/10.1016/j.ypmed.2011.12.020
        • Jans M.P.
        • Proper K.I.
        • Hildebrandt V.H.
        Sedentary behavior in Dutch workers: differences between occupations and business sectors.
        Am J Prev Med. 2007; 33: 450-454https://doi.org/10.1016/j.amepre.2007.07.033
        • Sugiyama T.
        • Ding D.
        • Owen N.
        Commuting by car: weight gain among physically active adults.
        Am J Prev Med. 2013; 44: 169-173https://doi.org/10.1016/j.amepre.2012.09.063
        • Tamers S.L.
        • Okechukwu C.
        • Marino M.
        • Guéguen A.
        • Goldberg M.
        • Zins M.
        Effect of stressful life events on changes in smoking among the French: longitudinal findings from GAZEL.
        Eur J Public Health. 2015; 25: 711-715https://doi.org/10.1093/eurpub/ckv036
        • McKee S.A.
        • Maciejewski P.K.
        • Falba T.
        • Mazure C.M.
        Sex differences in the effects of stressful life events on changes in smoking status.
        Addiction. 2003; 98: 847-855https://doi.org/10.1046/j.1360-0443.2003.00408.x
        • Mehra R.
        Retire for better sleep?.
        Sleep. 2009; 32: 1410-1411
        • Kivimäki M.
        • Nyberg S.T.
        • Batty G.D.
        • et al.
        Job strain as a risk factor for coronary heart disease: a collaborative meta-analysis of individual participant data.
        Lancet. 2012; 380: 1491-1497https://doi.org/10.1016/S0140-6736(12)60994-5
        • Kivimäki M.
        • Virtanen M.
        • Elovainio M.
        • Kouvonen A.
        • Väänänen A.
        • Vahtera J.
        Work stress in the etiology of coronary heart disease—a meta-analysis.
        Scand J Work Environ Health. 2006; 32: 431-442https://doi.org/10.5271/sjweh.1049
        • Barnett I.
        • van Sluijs E.M.
        • Ogilvie D.
        Physical activity and transitioning to retirement: a systematic review.
        Am J Prev Med. 2012; 43: 329-336https://doi.org/10.1016/j.amepre.2012.05.026
        • Chung S.
        • Domino ME
        • Stearns SC
        • Popkin BM
        Retirement and physical activity: analyses by occupation and wealth.
        Am J Prev Med. 2009; 36: 422-428https://doi.org/10.1016/j.amepre.2009.01.026
        • Jokela M.
        • Ferrie J.E.
        • Gimeno D.
        • et al.
        From midlife to early old age: Health trajectories associated with retirement.
        Epidemiology. 2010; 21: 284-290https://doi.org/10.1097/EDE.0b013e3181d61f53
        • Mealing N.
        • Banks E.
        • Jorm L.
        • Steel D.
        • Clements M.
        • Rogers K.
        Investigation of relative risk estimates from studies of the same population with contrasting response rates and designs.
        BMC Med Res Methodol. 2010; 10: 26https://doi.org/10.1186/1471-2288-10-26